Spline technologies in the study of the key risk factors of forecast quality for winter wheat yield

Abstract

Purpose: the author researches the repeatability laws of the «secondary» feature (winter wheat’s yield) in relation to the «primary» features (solar activity, precipitation, ambient temperature) with the use of original methods of Fourier-synchronism of V.A. Tartakovskiy in this paper. In addition, the author determines the shares of these periodic interactions and uses it in the generation of multidimensional generalized forecast. Discussion: the author developed and offered continuous spline polyforme piecewise approximation models for the analysis and forecast of the yield of grain, search, visualization, mathematical description and calculation of cycles. These models are different from the known in the literature momomodels and the least squares method analytic and suitability to systems with variable structure. Besides, the author considers the results in phase space, defines the topological and metrical characteristics of the cycles (start, end, length, scale variable and its first derivative, etc.). Results: the author adapted and improved the complex of socio-economic mathematical models subject to risk factors. This complex is the main component of the multicriteria approach with direct decision support methods for investment safety (definition and risk reduction) in agriculture.